Apple is reportedly training a large language model specifically for the Chinese market with technical support from Alibaba, according to multiple outlets citing people familiar with the effort. The move would run alongside Apple’s earlier plan to use Alibaba’s Qwen model for parts of Apple Intelligence in mainland China. Apple has not announced a timetable or confirmed details publicly.
The Reuters-based reports say the arrangement is enabled by China’s regulatory approval process. China’s Cyberspace Administration registered Apple’s generative AI service in July, clearing a key hurdle that had delayed Apple Intelligence on-device in mainland China. Under the approved framework, Alibaba’s Qwen is described as integrating into Apple Intelligence across compatible devices sold in China, while Baidu is mentioned as supporting search-related functions.
Outlets characterize the reported custom model as a departure from Apple’s prior approach of relying primarily on third-party models for regional generative AI. Some reports add that OpenAI’s ChatGPT and Anthropic’s Claude are unavailable in China, while Apple’s exact task split between the proprietary model, Qwen, and Baidu has not been disclosed. Several reports also note that Apple Intelligence is expected to launch in China after an iOS update, but no specific launch date is provided.